Practice exercises and activities based on the book "Interactive Data Visualization with Python: Present your data as an effective and compelling story," 2nd Edition

Book link: https://www.amazon.com/Interactive-Data-Visualization-Python-compelling/dp/1800200943

Date: September 2021

Author: Steven Ponce

Chapter 3: From Static to Interactive Visualization

# Load libraries

import sys
import math
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt

# ignore warnings
import warnings
warnings.filterwarnings("ignore")

print('You\'re running python %s' % sys.version.split(' ')[0])

You're running python 3.8.3

from bokeh.resources import INLINE import bokeh.io from bokeh import * bokeh.io.output_notebook(INLINE)

# loading CO2 data
df_co2 = pd.read_csv('co2.csv')
df_co2.tail()

	country	1800	1801	1802	1803	1804	1805	1806	1807	1808	...	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014
187	Venezuela	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	6.160	6.220	5.810	6.360	6.290	6.510	6.000	6.650	6.070	6.030
188	Vietnam	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	1.160	1.210	1.220	1.360	1.470	1.610	1.700	1.570	1.610	1.800
189	Yemen	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.974	1.010	0.964	0.999	1.070	0.993	0.811	0.749	0.997	0.865
190	Zambia	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.190	0.185	0.152	0.166	0.186	0.194	0.206	0.249	0.261	0.288
191	Zimbabwe	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.832	0.796	0.742	0.573	0.406	0.552	0.665	0.530	0.776	0.780

5 rows × 216 columns

# loading gapminder data
gm = pd.read_csv('gm.csv', encoding=('ISO-8859-1'))
gm.tail()

	Country	Year	fertility	life	population	child_mortality	gdp	region
10106	Åland	2002	NaN	81.80	26257.0	NaN	NaN	Europe & Central Asia
10107	Åland	2003	NaN	80.63	26347.0	NaN	NaN	Europe & Central Asia
10108	Åland	2004	NaN	79.88	26530.0	NaN	NaN	Europe & Central Asia
10109	Åland	2005	NaN	80.00	26766.0	NaN	NaN	Europe & Central Asia
10110	Åland	2006	NaN	80.10	26923.0	NaN	NaN	Europe & Central Asia

gm.shape

(10111, 8)

# drop duplicates
df_gm = gm[['Country', 'region']].drop_duplicates()
df_gm.tail()

	Country	region
9901	Western Sahara	Middle East & North Africa
9951	Yemen, Rep.	Middle East & North Africa
10001	Zambia	Sub-Saharan Africa
10051	Zimbabwe	Sub-Saharan Africa
10101	Åland	Europe & Central Asia

df_gm.shape

(204, 2)

# merge the two DF
'''
You can use merge() any time when you want to do database-like join operations
'''
df_w_regions = pd.merge(df_co2, df_gm, left_on='country', right_on='Country', how='inner')
df_w_regions.tail()

	country	1800	1801	1802	1803	1804	1805	1806	1807	1808	...	2007	2008	2009	2010	2011	2012	2013	2014	Country	region
168	Vanuatu	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.450	0.423	0.524	0.512	0.546	0.459	0.420	0.595	Vanuatu	East Asia & Pacific
169	Venezuela	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	5.810	6.360	6.290	6.510	6.000	6.650	6.070	6.030	Venezuela	America
170	Vietnam	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	1.220	1.360	1.470	1.610	1.700	1.570	1.610	1.800	Vietnam	East Asia & Pacific
171	Zambia	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.152	0.166	0.186	0.194	0.206	0.249	0.261	0.288	Zambia	Sub-Saharan Africa
172	Zimbabwe	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.742	0.573	0.406	0.552	0.665	0.530	0.776	0.780	Zimbabwe	Sub-Saharan Africa

5 rows × 218 columns

df_w_regions.shape

(173, 218)

# drop one of the country column
df_w_regions = df_w_regions.drop('Country', axis='columns')
df_w_regions.tail()

	country	1800	1801	1802	1803	1804	1805	1806	1807	1808	...	2006	2007	2008	2009	2010	2011	2012	2013	2014	region
168	Vanuatu	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.222	0.450	0.423	0.524	0.512	0.546	0.459	0.420	0.595	East Asia & Pacific
169	Venezuela	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	6.220	5.810	6.360	6.290	6.510	6.000	6.650	6.070	6.030	America
170	Vietnam	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	1.210	1.220	1.360	1.470	1.610	1.700	1.570	1.610	1.800	East Asia & Pacific
171	Zambia	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.185	0.152	0.166	0.186	0.194	0.206	0.249	0.261	0.288	Sub-Saharan Africa
172	Zimbabwe	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	0.796	0.742	0.573	0.406	0.552	0.665	0.530	0.776	0.780	Sub-Saharan Africa

5 rows × 217 columns

df_w_regions.shape

(173, 217)

'''
Pandas.melt() unpivots a DataFrame from wide format to long format.
melt() function is useful to massage a DataFrame into a format where one or more columns are identifier variables,
while all other columns, considered measured variables, are unpivoted to the row axis, leaving just two non-identifier 
columns, variable and value.
'''
# melt
new_co2 = pd.melt(df_w_regions, id_vars=['country', 'region'])
columns = ['country', 'region', 'year', 'co2']
new_co2.columns = columns
new_co2.tail()

	country	region	year	co2
37190	Vanuatu	East Asia & Pacific	2014	0.595
37191	Venezuela	America	2014	6.030
37192	Vietnam	East Asia & Pacific	2014	1.800
37193	Zambia	Sub-Saharan Africa	2014	0.288
37194	Zimbabwe	Sub-Saharan Africa	2014	0.780

new_co2.shape 

(37195, 4)

# set 1964 and onward as the range for year and int64 as data type.
# sort DF by country and then year
df_co2 = new_co2[new_co2['year'].astype('int64') > 1963]
df_co2['year'] = df_co2['year'].astype('int64')
df_co2 = df_co2.sort_values(by=['country', 'year'])
df_co2.head()

	country	region	year	co2
28372	Afghanistan	South Asia	1964	0.0863
28545	Afghanistan	South Asia	1965	0.1010
28718	Afghanistan	South Asia	1966	0.1080
28891	Afghanistan	South Asia	1967	0.1240
29064	Afghanistan	South Asia	1968	0.1160

# create new DF called df_gdp
df_gdp = gm[['Country', 'Year','gdp']]
df_gdp.columns = ['country', 'year','gdp']
df_gdp.head()

	country	year	gdp
0	Afghanistan	1964	1182.0
1	Afghanistan	1965	1182.0
2	Afghanistan	1966	1168.0
3	Afghanistan	1967	1173.0
4	Afghanistan	1968	1187.0

df_gdp.shape

(10111, 3)

# Now we have two DF for CO2 and GDP
# Lets merge them together
data = pd.merge(df_co2, df_gdp, on=['country', 'year'], how='left')
data = data.dropna()
data.head()

	country	region	year	co2	gdp
0	Afghanistan	South Asia	1964	0.0863	1182.0
1	Afghanistan	South Asia	1965	0.1010	1182.0
2	Afghanistan	South Asia	1966	0.1080	1168.0
3	Afghanistan	South Asia	1967	0.1240	1173.0
4	Afghanistan	South Asia	1968	0.1160	1187.0

data.shape

(8202, 5)

# create a numoy array of the co2 and gdp columnms
np_co2 = np.array(data['co2'])
np_gdp = np.array(data['gdp'])

# correlation between co2 and gpd
''''Return Pearson product-moment correlation coefficients.'''
corr = np.corrcoef(np_co2, np_gdp)
corr[1,0]

0.7821973088777349

Excercise creating the base static plot for an interactive viz

# import libraries
from bokeh.io import curdoc, output_notebook
from bokeh.plotting import figure, show
from bokeh.models import HoverTool, ColumnDataSource, CategoricalColorMapper, Slider
from bokeh.palettes import Spectral6
from bokeh.layouts import widgetbox, row, column
output_notebook()

Loading BokehJS ...

# color code the datapoints
regions_list = data.region.unique().tolist()
regions_list

['South Asia',
 'Europe & Central Asia',
 'Middle East & North Africa',
 'Sub-Saharan Africa',
 'America',
 'East Asia & Pacific']

# assigning a color from spectral16
color_mapper = CategoricalColorMapper(factors=regions_list, palette=Spectral6)

# data source for the plot
source = ColumnDataSource(data={
    'x': data.gdp[data['year'] == 1964],
    'y': data.co2[data['year'] == 1964],
    'country': data.country[data['year'] == 1964],
    'region': data.region[data['year'] == 1964],
})

# store the min and max GDP values
xmin, xmax = min(data.gdp), max(data.gdp)

# store the min and max co2 values
ymin, ymax = min(data.co2), max(data.co2)

# Create the figure: plot
plot = figure(title='CO2 Emissions vs GDP in 1964', 
              plot_height=600, plot_width=950,
              x_range=(xmin, xmax),
              y_range=(ymin, ymax), y_axis_type='log')

# add cicular glyphs to the plot
plot.circle(x='x',
             y='y',
             fill_alpha=0.9,
             source=source,
             legend_group='region',
             color=dict(field='region', transform=color_mapper),
             size=10)

GlyphRenderer(

id = '1041', …)

data_source = ColumnDataSource(id='1004', ...),

glyph = Circle(id='1039', ...),

hover_glyph = None,

js_event_callbacks = {},

js_property_callbacks = {},

level = 'glyph',

muted = False,

muted_glyph = None,

name = None,

nonselection_glyph = Circle(id='1040', ...),

selection_glyph = 'auto',

subscribed_events = [],

syncable = True,

tags = [],

view = CDSView(id='1042', ...),

visible = True,

x_range_name = 'default',

y_range_name = 'default')

# Set the legend.location attribute of the plot
plot.legend.location = 'bottom_right'

# Set the x-axis label
plot.xaxis.axis_label = 'Income Per Person'

# Set the y-axis label
plot.yaxis.axis_label = 'CO2 Emissions (tons per person)'

from bokeh.models import PrintfTickFormatter
plot.xaxis[0].formatter = PrintfTickFormatter(format="%1f")

show(plot)

# adding a slider

start = min(data.year)
end = max(data.year)
value = min(data.year)

slider = Slider(start=start, end=end, step=1, value=value, title='Year')

# Define the callback: update_plot
def update_plot(attr, old, new):
    
    # set the `yr` name to `slider.value` and `source.data = new_data`
    yr = slider.value
    
    new_data = {
    'x': data.gdp[data['year'] == yr],
    'y': data.co2[data['year'] == yr],
    'country': data.country[data['year'] == yr],
    'region': data.region[data['year'] == yr],
    }
    
    source.data = new_data
    
    # Add title to figure: plot.title.tex
    plot.title.text = 'CO2 Emissios vs. GDP in %d' %yr
    

data.keys()

Index(['country', 'region', 'year', 'co2', 'gdp'], dtype='object')

# Attach the callback to the 'value' property of slider
slider.on_change('value', update_plot)

# layout = row(widgetbox(slider), plot)       #### widgetbox is deprecated
layout = column(slider)

curdoc().add_root(layout)
show(layout)

WARNING:bokeh.embed.util:
You are generating standalone HTML/JS output, but trying to use real Python
callbacks (i.e. with on_change or on_event). This combination cannot work.

Only JavaScript callbacks may be used with standalone output. For more
information on JavaScript callbacks with Bokeh, see:

    https://docs.bokeh.org/en/latest/docs/user_guide/interaction/callbacks.html

Alternatively, to use real Python callbacks, a Bokeh server application may
be used. For more information on building and running Bokeh applications, see:

    https://docs.bokeh.org/en/latest/docs/user_guide/server.html

# adding a hoover tool
hover = HoverTool(tooltips=[('Country', '@country'), ('GDP', '@x'),
                           ('CO2 Emission', '@y')])

plot.add_tools(hover)
show(layout)
show(plot)

WARNING:bokeh.embed.util:
You are generating standalone HTML/JS output, but trying to use real Python
callbacks (i.e. with on_change or on_event). This combination cannot work.

Only JavaScript callbacks may be used with standalone output. For more
information on JavaScript callbacks with Bokeh, see:

    https://docs.bokeh.org/en/latest/docs/user_guide/interaction/callbacks.html

Alternatively, to use real Python callbacks, a Bokeh server application may
be used. For more information on building and running Bokeh applications, see:

    https://docs.bokeh.org/en/latest/docs/user_guide/server.html

Interactive visualization w/ Plotly Express

import pandas as pd
import plotly.express as px

# CO2 and GDP data
data.head()

	country	region	year	co2	gdp
0	Afghanistan	South Asia	1964	0.0863	1182.0
1	Afghanistan	South Asia	1965	0.1010	1182.0
2	Afghanistan	South Asia	1966	0.1080	1168.0
3	Afghanistan	South Asia	1967	0.1240	1173.0
4	Afghanistan	South Asia	1968	0.1160	1187.0

# store the min and max GDP values
xmin, xmax = min(data.gdp), max(data.gdp)

# store the min and max co2 values
ymin, ymax = min(data.co2), max(data.co2)

# scatter plot
fig = px.scatter(data, 
                x='gdp', 
                y='co2',
                animation_frame='year',
                 animation_group='country',
                 color='region',
                 hover_name='country',
                 facet_col='region',
                 width=1550, height=400,
                 log_x=True,
                 size_max=45,
                 range_x=[xmin, xmax],
                 range_y=[ymin, ymax]
                )

fig.show()

Activity 3: Creating different interactive visualization using plotly express

Step 1 - recreate CO2 and GDP DF

import pandas as pd
import plotly.express as px

# CO2 and GDP data from previous excercise
data.tail()

	country	region	year	co2	gdp
8817	Zimbabwe	Sub-Saharan Africa	2009	0.406	1352.0
8818	Zimbabwe	Sub-Saharan Africa	2010	0.552	1484.0
8819	Zimbabwe	Sub-Saharan Africa	2011	0.665	1626.0
8820	Zimbabwe	Sub-Saharan Africa	2012	0.530	1750.0
8821	Zimbabwe	Sub-Saharan Africa	2013	0.776	1773.0

Step 2 - create a scatter plot with x- and y-axes as year and co2 respectively. Add box plot for the co2 values with the marginal_y parameter

import plotly.express as px

# scatter plot & box plot using px
fig1 = px.scatter(data, x=data.year, 
                  y=data.co2, 
                  color=data.region, 
                  marginal_y='box')              
fig1.show()

Step 3 - create a rug plot for gdp values with the marginal_x parameter. Add the animation on the year column

# rug plot
fig2 = px.scatter(data, 
                  x=data.gdp, 
                  y=data.co2, 
                  color=data.region,
                  marginal_x='rug',
                  marginal_y='box',
                  animation_frame=data.year,
                  animation_group=data.country,
                 )              
fig2.show()

Step 4 - create a density contour with the x- and y-axes as gdp and co2.

# density contour plot
fig3 = px.density_contour(data, 
                          x=data.gdp, 
                          y=data.co2, 
                          color=data.region,
                          marginal_x='rug',
                          marginal_y='box',
                          animation_frame=data.year,
                          animation_group=data.country,
                         )
fig3.show()

data.keys()

Index(['country', 'region', 'year', 'co2', 'gdp'], dtype='object')